Seems reasonable.
Do they track the productivity and egg quality of individual hens? The laggards could be sent to become McNuggets.
-- John Larkin Highland Technology, Inc jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
They would be the democrats.
Yes, chickens and turkeys.
Seriously, Democrats seem to be a lot more afraid of things than Rs or independents. More self-obcessed and neurotic, too. The New York Times is an appalling mass of neurosis and narcissism and bitterness.
-- John Larkin Highland Technology, Inc jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Wow, did you get the medical certs on that too? LOL
Tim
-- Seven Transistor Labs Electrical Engineering Consultation Website: http://seventransistorlabs.com
I seriously doubt Hobbs has done independent research on the spectral absor ption characteristics of eggs. He's obviously bouncing off prior research o n the subject by an agricultural scientist. But this information, and all t he electronics minutiae, can't be intelligently evaluated without alpha-beta type of design requirement: acceptable numbers of false rejects and false accepts.
On a sunny day (Sat, 08 Nov 2014 14:37:42 -0500) it happened Phil Hobbs wrote in :
I have seen size sorting machines that had a shaking place with holes in it, what was smaller dropped through. Maybe not so good for eggs ... How do you measure size, with fotocells? How about using a simple color camera for blood spots and cracks? Hey I think I can do that. Look for the UV in the YUV
30 eggs per second.
Na, in the US they hit 2 or 3 out of 4 IIRC. Yet they get billions funding for the anti anti anti (forget how many antis) missile system.
People eat a lot of eggs, and of course lots and lots of work has been being done on automatic egg grading systems and many, many other aspects of the egg production and egg-breaking industries. ("Egg breaking" really is the name of the industry that produces egg products such as scrambled egg mix and egg ingredients for cakes and so forth.) I've done enough of my own measurements to be sure that the instrument will work satisfactorily over the full range of egg types I can find in my local supermarkets.
The customer was originally going to arrange to have local egg farms send me some blood eggs to test, but decided that the dye calibration procedure was easier and more repeatable anyway.
The topic of the thread is not whether whoever goes by the pseud "Fred Bloggs" (which is sort of the English equivalent of "John Doe" or "Larry Lunchbucket") is satisfied with the SPC aspects of the instrument. Sorry about that.
Of course, if you want to go into the egg sorting machine business, we should talk. ;)
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Do they *make* Six Sigma black belts in sizes to fit chickens? :)
If somebody wanted to implement JIT, I wonder how big a chicken coop you'd need to take care of the average Waffle House/Denny's/IHOP/etc. Also, you'd have to train the hens to read the kanban board.
Matt Roberds
I got all that were required for a research instrument. The gizmo worked fine, although the uncertainty in the grating position was a limitation. It needed an auxiliary optical sensor to look at the specular reflection from the grating, with a lateral effect cell or something like that. (I'll put that in once the startup gets a bit more dough--it's only a couple of days' worth of work.)
(You gotta watch that 'LOL' stuff, Tim, it makes you look like a
12-year-old.)Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
On Sun, 9 Nov 2014 18:16:03 +0000 (UTC), snipped-for-privacy@att.net Gave us:
Some of those MPCs that are touted as being so correct are only correct in certain manufacturing environments.
LEAN and SIGMA6, 5S and all that happy too-many-chiefs horseshit are meant for automated process refinement in large, high volume demand settings.
It fails miserably when a single person is responsible for building an entire CCI mil spec transceiver.
There is no push line for that, and it is an insult to come into a setting and tell someone that has been doing it for years that they "need to clean their work area up", or that that environmental aspect of the production floor will make things more efficient.
In certain setting all it does is waste millions while some asshole sits and blows his little horn, claiming gains of some sort after spending millions on implementation. It will NEVER amortize its cost, and again, is an absolute insult to what our culture used to be before the retarded "You work for the empire" indentured slave mentality set in. But folks don't care, because the pay is nice, right?
LOL!
( adults are just tall children )
-- Les Cargill
On Sun, 09 Nov 2014 13:24:51 -0600, Les Cargill Gave us:
Or fat...
BRL! (Belly Roll Laugh!)
Or inebriated...
BRL!! (Bonghit Release Laugh!!)
LOL is so lame '90s.
Hell it is worse than disco or polyester shirts with pirate sleeves!
Your customer and especially his customers are not going to be particularly impressed with the technology inside your little grading machine if it cos ts them profits. The alpha-beta characteristic performance should be the st arting point because that's what the users are most concerned about. From t hat you develop an error budget for the inevitably imperfect final product, said error budget to be distributed across the performance of the sub-syst ems in accordance with known sensitivities obtained by measurement or desig n. You have done none of that, and that's why you keep evading the topic.
Hey, 96% on the BlatherOmeter!
-- John Larkin Highland Technology, Inc jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
"Evading the topic", hmm? You know, Fred, where I come from we know a trick worth two of that one.
You're on much more promising territory when you suggest running SPC on _hens_. Of course, you'll have to do extensive interviews with a representative sample of the world's 20 billion hens to find out what incentives they might need to produce more eggs and what sort of care might lead them to bleed less while doing so. Your unique brand of compassion would be just the ticket.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
I think that they generally use suction cups spaced at decreasing distances from the eggs as they pass by. Extra large eggs are the ones picked up by the first row of suction cups, large eggs by the second, and so on.
Cracks are detected using acoustic methods. I haven't gone into it at all, but I expect that you jiggle the egg at some frequency and look for the second harmonic in the response. You can push on a crack, but you can't pull, so there would be some amount of rectification of the acoustic signal that you could see pretty easily, I expect.
Cameras aren't as suitable, because (a) you can only inspect ~40% of the surface area per camera, and (b) the eggs are whizzing by on conveyors, which partially obstruct the view.
Hard to do with your typical 8 bits per colour, even if the colours were chosen correctly, which they aren't. That's the typical software guy's approach to anything optical--point a camera at it, and crunch the daylights out of the resulting very low quality data. It works once in awhile, on sufficiently easy problems.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
ts
ry
rly impressed with the technology inside your little grading machine if it costs them profits. The alpha-beta characteristic performance should be the starting point because that's what the users are most concerned about. Fro m that you develop an error budget for the inevitably imperfect final produ ct, said error budget to be distributed across the performance of the sub-s ystems in accordance with known sensitivities obtained by measurement or de sign. You have done none of that, and that's why you keep evading the topic .
Is that mechanics' humor?
-Lasse
You're clueless about eggs. Those things vary all over the place. You can even tell the season from their composition.
The problem is that there isn't that much light available, after passing twice through the shell of a very brown egg. That means a large photodiode (1 cm**2), which means a lot of capacitance.
Even with an efficient bootstrap, that capacitance leads to a noise contribution that rises steeply with frequency, so slower pulsing is better. OTOH the measurement time available is only about 1/20 of a second, which isn't a long time for a lock-in measurement. Thus I really need to switch the LED somewhere in the 1-2 kHz range to avoid losing performance.
I could do that using a big fat inductor, but I'd have to get rid of the ripple by dumping current. Given the tolerances (22-26V on the supply,
25-29V for the LED), doing that the simple way would be pretty wasteful compared with a boost regulator and a couple of 3300 uF caps.A 900 kHz switching regulator with a couple of stages of RC filtering on the front (0.27 ohms and 3300 uF) seems to do a pretty good job with the
1 kHz stuff, and a couple of LC stages gets rid of the MF conducted crap down to a pretty low level. We'll see.Charge pumps are pretty lossy unless the caps are really big, and if they are, in the 20W range the peak switch currents become a bit of a problem.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.